Re: [PATCH] common/sfc_efx/base: remove VQ index check during VQ start

[Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>]

On 7/8/22 10:37, abhimanyu.saini@xilinx.com wrote:
> From: Abhimanyu Saini <absaini@amd.com>
> 
> The used/avail queue indexes are not bound by queue
> size, because the descriptor entry index is calculated
> by a simple modulo between queue index and queue_size

"is calculated" is a bit vague since looking at the code
I've failed to find the place where modulo operation is
done. Don't we need to apply it these values are put
into MCDI message?

> 
> So, do not check initial used and avail queue indexes
> against queue size because it is possible for these
> indexes to be greater than queue size in the
> following cases:
> 1) The queue is created to be migrated into, or
> 2) The client issues a qstop/qstart after running datapath
> 
> Fixes: 4dda72dbdeab3 ("common/sfc_efx/base: add base virtio support for vDPA")
> Cc: stable@dpdk.org
> 
> Signed-off-by: Abhimanyu Saini <absaini@amd.com>
> ---
>   drivers/common/sfc_efx/base/rhead_virtio.c | 12 +-----------
>   1 file changed, 1 insertion(+), 11 deletions(-)
> 
> diff --git a/drivers/common/sfc_efx/base/rhead_virtio.c b/drivers/common/sfc_efx/base/rhead_virtio.c
> index 335cb74..7f08717 100644
> --- a/drivers/common/sfc_efx/base/rhead_virtio.c
> +++ b/drivers/common/sfc_efx/base/rhead_virtio.c
> @@ -47,14 +47,6 @@
>   		goto fail2;
>   	}
>   
> -	if (evvdp != NULL) {
> -		if ((evvdp->evvd_vq_cidx > evvcp->evvc_vq_size) ||
> -		    (evvdp->evvd_vq_pidx > evvcp->evvc_vq_size)) {
> -			rc = EINVAL;
> -			goto fail3;
> -		}
> -	}
> -
>   	req.emr_cmd = MC_CMD_VIRTIO_INIT_QUEUE;
>   	req.emr_in_buf = payload;
>   	req.emr_in_length = MC_CMD_VIRTIO_INIT_QUEUE_REQ_LEN;
> @@ -116,15 +108,13 @@
>   
>   	if (req.emr_rc != 0) {
>   		rc = req.emr_rc;
> -		goto fail4;
> +		goto fail3;
>   	}
>   
>   	evvp->evv_vi_index = vi_index;
>   
>   	return (0);
>   
> -fail4:
> -	EFSYS_PROBE(fail4);
>   fail3:
>   	EFSYS_PROBE(fail3);
>   fail2: